This paper proposes a bilevel programming model for the planning of tidal current farms (TCFs). The micrositing strategy of tidal current turbines (TCTs) and the collector system planning scheme are coordinated to achieve a better balance of energy yields, TCF capital investment and power systems economic operation using the proposed method. The power output of the TCF is modeled considering the characteristics of tidal current velocity and wake effects. A coordinated planning model consisting of one upper level model and two lower level models is developed to maximize comprehensive profit. Not only the investment and maintenance costs of TCTs and submarine cables, but also the operation cost of the collector system and the impact of the TCF integration on the operation of power systems are all taken into account to ensure the long-term benefits of both TCF owners and power systems. An efficient solution for the proposed planning model is developed by combining a genetic algorithm with a mixed integer programming. The effectiveness and adaptability of the proposed method are demonstrated using the measured data of tidal current velocity with distinct characteristics and the IEEE 30- and 118-bus test systems.